Long-Term Return Reversals: Overreaction or Taxes?




    Search for more papers by this author
    • George is from C.T. Bauer College of Business, University of Houston. Hwang is from Nanyang Business School, Nanyang Technological University. We are grateful to Jonathan Berk, Alon Brav, Mark Grinblatt, Tom Rourke, Robert Stambaugh (the editor), an anonymous referee, and seminar participants at Hong Kong University of Science and Technology and the 2006 WFA meetings for helpful discussions and comments. Harry Leung provided excellent research assistance. George acknowledges financial support of the C.T. Bauer Professorship. Hwang acknowledges RGC grant HKUST6011/00H.


Long-term reversals in U.S. stock returns are better explained as the rational reactions of investors to locked-in capital gains than an irrational overreaction to news. Predictors of returns based on the overreaction hypothesis have no power, while those that measure locked-in capital gains do, completely subsuming past returns measures that are traditionally used to predict long-term returns. In data from Hong Kong, where investment income is not taxed, reversals are nonexistent, and returns are not forecastable either by traditional measures or by measures based on the capital gains lock-in hypothesis that successfully predict U.S. returns.